Menlo Ventures just closed $3 billion in fresh capital—the biggest fundraise in the firm's 50-year history. The headline number is impressive, but the real story is how a single audacious bet on Anthropic transformed a mid-tier VC shop into one of the most credible AI-focused investors on the planet.

The Bet That Could Have Gone Very Wrong

Rewind to 2024. The post-pandemic VC hangover was still very much in effect. SoftBank and Tiger Global were quietly nursing their wounds from a few years of writing checks like drunk sailors. Nobody was handing out $750 million for a single deal. Nobody, that is, except Menlo.

Menlo preemptively led Anthropic's Series D—a $750 million round that pushed the AI safety-obsessed model maker's valuation to $18.4 billion. That's a 4x step-up on the previous round, for those keeping score. The firm had been in since the Series C, back when Anthropic was more manifesto than product. By the time the Series D closed, Claude existed, Amazon had committed $4 billion to the cause, and the founding team—ex-OpenAI researchers Dario and Daniela Amodei—had serious credibility. So the bet itself wasn't insane. What was eyebrow-raising was how Menlo actually wrote the check.

The SPV Mechanic That Made It Possible

Of the $750 million total, roughly $500 million came through a Special Purpose Vehicle—a purpose-built investment entity that pools capital from multiple limited partners for a single deal. Menlo's own fund and insiders contributed the remaining $250 million. At the time, this was a creative workaround. Today, AI SPVs have become so ubiquitous they're practically a nuisance—Anthropic itself had to publicly call out unauthorized SPVs and secondary markets peddling fake access to its shares as outright scams. The structure Menlo pioneered has since been copied, watered down, and monetized into oblivion by everyone with a Substack and a wire transfer account.

But for the people who got in through Menlo's original, above-board vehicle? Their Anthropic stake is now worth roughly $14 billion. That's the kind of return that makes LPs forget every quarterly meeting where you had to explain why valuations were "temporarily compressed."

Anthology: The Sleeper Play Nobody Talked About

Menlo didn't just ride Anthropic's coattails—they co-launched a $100 million startup fund with Anthropic called Anthology. Cute name aside, the thing has quietly grown to $250 million in deployed capital, backed more than 60 companies, and already generated exits. Graphite got acquired by Cursor. Astrix Security landed at Cisco. These aren't vanity acquisitions; they're real liquidity events from a fund that's barely a year old.

The strategic logic here is clever: by co-investing alongside Anthropic, Menlo gets a ground-floor view of what Claude can actually do in production environments, which startups are building something real versus wrapping an API call in a pretty UI, and which application layers have genuine defensibility. That's market intelligence you can't buy from a research report.

The Portfolio That Followed

Off the back of the Anthropic relationship and the Anthology network, Menlo has assembled a roster of AI companies that reads like a who's-who of the current cycle: OpenRouter, Higgsfield, Legora, Lovable, OpenEvidence. Whether all of these pan out is a separate question—plenty of today's AI darlings are burning capital at rates that would make a 2021-era crypto bro blush—but the point is that Menlo now has deal flow and pattern recognition that most VCs would pay for.

What the $3B Actually Signals

A $3 billion fund is also a $3 billion obligation. LPs are expecting Menlo to deploy that capital into opportunities as good as or better than Anthropic at a $18 billion valuation. That's a tall order when foundation model companies are now raising at valuations that would have been considered absurd two years ago, and the application layer is crowded with companies that are one GPT-5 update away from being disrupted.

Menlo's Anthropic bet worked because they moved early, structured creatively, and got lucky with timing. The next $3 billion will require them to do that again—repeatedly, across a much noisier landscape. History is littered with VCs who had one great fund followed by a spectacular regression to the mean.

But credit where it's due: they saw the wave early, found a way to surf it when no one else would, and built real infrastructure around the initial bet. That's not hype. That's execution.